119 – Statistical Data Editing Modernization
Evaluating Imputation Methods for the Agricultural Resource Management Survey
Andrew Dau
U.S. Department of Agriculture
Darcy Miller
National Agricultural Statistics Service
Audra Zakzeski
U.S. Department of Agriculture
The National Agricultural Statistics Service (NASS), in conjunction with the Economic Research Service (ERS), conducts the three-phase Agricultural Resource Management Survey (ARMS) to study the economic well-being of farm households. Since 2015, Iterative Sequential Regression (ISR), a multivariate imputation methodology, has been used to address item nonresponse in the third phase of the survey (ARMS 3). ISR is an in-house developed software program that requires a significant amount of support to maintain. Also, ISR was developed for use on continuous and semi-continuous data, and NASS wants to impute other data types, including categorical and ordinal data. Hence, NASS is exploring alternative “off-the-shelf” imputation approaches, specifically, IVEware, a product of the University of Michigan, and the Fully Conditional Specification Option in SAS® PROC MI. A 2018 JSM paper empirically compared ISR to these two alternatives using a subset of ARMS 3 data. This paper builds on that simulation work and culminates in an impact assessment of a change to one of the alternatives on reported estimates and operational resources through an application to the full ARMS 3 dataset.